Abstract

This meta-analysis examined the relationship between student engagement and mathematics achievement among nontraditional students, a population often balancing academic demands with work and family responsibilities. A systematic search of academic databases yielded 7,826 articles, from which four studies met the inclusion criteria. While only one study directly measured mathematics achievement, the others assessed academic achievement through GPA, which served as a proxy in the absence of mathematics achievement-specific data. Using a random-effects model and Fisher r-to-z transformed correlation coefficients, the overall effect size was found to be statistically significant (r = 0.42, p = .007), indicating a moderate positive association between engagement and mathematics achievement. Engagement dimensions included behavioral, cognitive, and emotional aspects. Substantial heterogeneity was detected (I² = 92.86%), suggesting meaningful variation in study contexts, engagement measures, and sample characteristics. No publication bias or influential outliers were identified. Findings highlight the importance of fostering student engagement to support mathematics achievement or academic achievement in general among nontraditional learners. While the limited number of mathematics-specific studies presents a constraint, the results suggest that engagement – including constructs like math self-efficacy – can positively influence achievement. Further research is needed to directly investigate how different engagement types affect mathematics achievement of nontraditional students.

1. Introduction

It is well-documented across educational research that student engagement plays a crucial role in shaping academic achievement, particularly in the domain of mathematics ^{1, 2}. Mathematics is a foundational subject that often poses significant challenges to learners, and sustained engagement – whether behavioral, cognitive, or emotional – has been associated with improved performance, persistence, and positive attitudes in math learning contexts ^{3, 4}. Student engagement is commonly defined and measured across three dimensions: behavioral engagement (e.g., attendance, participation, study habits), cognitive engagement (e.g., self- regulated learning, critical thinking), and emotional/social engagement (e.g., motivation, student-faculty interactions) ^{1, 5}.

Understanding the dynamics of engagement is especially important for nontraditional students, who often juggle academic responsibilities alongside work, family, and other adult commitments. For this population, maintaining engagement in mathematics can be both difficult and critical for academic success. Despite the growing body of literature on engagement and mathematics achievement, there remains limited synthesis on how this relationship manifests specifically among nontraditional students. This gap highlights the need for a meta-analytic review focused on clarifying the strength of the association between engagement and mathematics achievement within this unique student demographic ⁶.

Nontraditional students constitute a significant and growing portion of the higher education population, particularly within community colleges, where they enroll in large numbers ^{7, 8}. Compared to traditional students, nontraditional learners are typically older, financially independent, employed while studying, and often balancing family responsibilities ^{9, 10}. These demographic characteristics often shape their learning experiences, particularly in mathematics courses, which are frequently perceived as challenging and anxiety-inducing ¹¹. Due to time constraints, prior negative experiences, or gaps in foundational knowledge, nontraditional students may experience lower confidence and persistence in mathematics-related tasks ^{12, 13}. This presents unique challenges in fostering sustained engagement in mathematics learning environments. Engagement – defined as the level of participation in educational activities, motivation, and cognitive investment – has been widely recognized as a key predictor of mathematics achievement ^{14, 15}. Therefore, understanding how nontraditional students engage in mathematics is critical to supporting their academic progress in this domain.

Prior research has extensively examined student engagement and its role in learning outcomes, but much of the literature has focused on traditional students, leaving a gap in understanding regarding how engagement operates within nontraditional student populations ^{6, 9}. While some studies suggest that engagement significantly enhances mathematics achievement ^{14, 16}, others indicate that external factors such as work schedules, financial constraints, and family responsibilities may moderate or weaken this relationship ^{12, 13}. Additionally, inconsistencies in measurement approaches – ranging from standardized engagement scales to course-specific evaluations – have led to variability in reported effect sizes ^{15, 17}. Notably, despite the growing attention to engagement and mathematics achievement in the general student population, very few empirical studies have specifically examined this relationship among nontraditional students. In fact, in this current review, only one of the included articles examined mathematics self-efficacy as a predictor of mathematics achievement. Although self-efficacy is conceptually distinct from engagement, it is strongly related particularly to cognitive and emotional engagement, as students with higher math self-efficacy tend to exert greater effort, persist longer, and experience more motivation and confidence in mathematical tasks ^{18, 19}. Research also suggests that math self-efficacy can foster behavioral engagement, as students are more likely to participate actively in class and complete challenging tasks when they believe in their capabilities ^{20, 21}. This significant gap in literature underscores the need for a systematic synthesis of available evidence to determine the strength and direction of the engagement-mathematics achievement relationship in nontraditional learners.

To address this gap in the literature, the present meta-analysis aims to synthesize existing quantitative studies examining the relationship between student engagement and mathematics achievement specifically among nontraditional students. By aggregating effect sizes and analyzing patterns across studies, this review seeks to clarify the strength and consistency of this association within this unique population. The central research question guiding this meta-analysis is: What is the strength of the association between student engagement and mathematics achievement among nontraditional students? In doing so, this study contributes to a more nuanced understanding of how engagement influences achievement in mathematics for learners who face distinct academic and life challenges. The findings are expected to inform educators, researchers, and policymakers on how best to support nontraditional students through engagement-focused strategies that can enhance mathematics achievement.

2. Method

To ensure methodological rigor, only peer-reviewed studies reporting statistical measures of engagement and mathematics achievement or academic achievement within nontraditional student populations were included. Studies that lacked quantitative data, exclusively examined traditional students, or did not assess engagement as a predictor variable were excluded. This meta-analysis employs effect size calculations, heterogeneity analysis, and publication bias assessments to provide a more precise estimate of the relationship between engagement and mathematics or academic achievement.

This meta-analysis systematically investigated the relationship between student engagement and mathematics or general academic achievement among nontraditional students. To ensure transparency and reproducibility, the study followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.

The initial phase of the literature search focused on mathematics achievement specifically. Using the keywords "Relationship" AND "Engagement" AND "Mathematics Achievement" AND "Nontraditional Students," searches were conducted across Crossref (k = 1,000), Google Scholar (k = 200), and Open Lens (k = 250), yielding a total of 1,450 articles. These articles were screened based on titles, abstracts, and the availability of full-text articles. However, only one study was identified that explicitly examined the relationship between engagement and mathematics achievement in the context of nontraditional students.

Given the limited number of mathematics achievement-specific studies, a second phase of the literature search was conducted using broader keywords: "Engagement" AND "Academic Achievement" AND "Nontraditional Students." This yielded 576 articles from ERIC and 5,800 from Google Scholar, for a total of 6,376 articles. After screening titles, abstracts, and availability of full-text articles, 6,292 articles were excluded, resulting in 84 potentially relevant articles.

Combining both phases, a total of 85 articles were identified. After removing two duplicates, 83 full-text articles were assessed for eligibility. Of these, 79 were excluded based on the study's predefined inclusion and exclusion criteria. Ultimately, only four studies met all inclusion criteria and were included in the meta-analysis.

Although only one of the included studies explicitly examined mathematics achievement, the other three used general academic achievement, primarily operationalized through GPA. While "Academic Achievement" is a broader construct, it is accepted as a valid proxy for mathematics achievement in meta-analytic contexts, especially when studies include performance indicators that may implicitly or explicitly involve mathematics ^{22, 23}. This broader scope enabled the inclusion of relevant empirical studies, ensuring a sufficient number of qualified studies for meaningful synthesis while maintaining alignment with the meta-analysis objective.

The inclusion criteria for study selection were as follows: First, studies needed to focus specifically on nontraditional students, defined as individuals older than 24 years, employed full-time, responsible for dependents, returning to education after a gap, or enrolled part-time in a community college setting ^{9, 24}. Second, only empirical quantitative studies were included, particularly those reporting correlational, regression, or effect-size statistics such as Pearson's r, Cohen's d, or odds ratios ^{25, 26}. Studies employing experimental, quasi-experimental, or observational research designs were also eligible. Third, studies had to define and measure student engagement, including behavioral engagement (e.g., attendance, participation, study habits), cognitive engagement (e.g., self-regulated learning, critical thinking), and emotional/social engagement (e.g., motivation, student-faculty interactions) ⁵. Academic achievement was required to be measured using at least one quantifiable outcome, such as Grade Point Average (GPA), final course grades, standardized or non-standardized test scores, retention rates, or course completion ²⁷. To ensure the relevance of findings, only studies published from 2010 to 2025 were considered, with older studies included only if they were seminal contributions to the field. Only English-language full-text articles were included, and eligible publications were restricted to peer-reviewed journal articles, dissertations, theses, or conference proceedings ²⁸.

Studies were excluded if they focused exclusively on traditional students (ages 18-24, full-time, recent high school graduates) or were conducted in four-year universities instead of community colleges ²⁴, Qualitative-only studies, such as case studies, interviews, and thematic analyses, were excluded, as well as opinion pieces, theoretical papers, and literature reviews without original data ²⁹. Additionally, meta-analyses were excluded since they synthesize existing research rather than contribute new data. Studies were also excluded if they did not explicitly measure engagement using a validated instrument or if they lacked statistical measures of academic achievement ²⁶.

The data extraction process was conducted by a single researcher to ensure consistency in identifying key study characteristics. The extracted data included study authors, year of publication, sample size, engagement measures, academic achievement measures, statistical analyses, and effect sizes, following established meta-analysis guidelines ^{30, 31}. The primary effect size considered in this meta-analysis was Pearson's correlation coefficient (r), as it directly measures the strength and direction of the association between student engagement and academic achievement ³².

A structured data extraction sheet was used to systematically record relevant information from each study, following best practices for systematic reviews ³³. In cases where studies reported multiple effect sizes, the most comprehensive or representative correlation was selected ³⁰. Specifically, for one study (Article 4) that reported separate correlation values for behavioral, cognitive, and emotional engagement, an average of the three r values was computed to derive an overall engagement-achievement correlation, following recommendations for synthesizing multidimensional engagement data ¹. For another study (Article 3) that did not report a direct correlation coefficient but provided a structural equation model (SEM) output, the correlation was estimated using the regression coefficient and the reported standard deviations of the predictor (engagement) and the outcome (GPA), following standard conversion procedures. All extracted data were compiled in Microsoft Excel for organization and further analysis. Studies that did not report Pearson's correlation coefficient (r) or did not provide sufficient statistical information for conversion were excluded from the meta-analysis to ensure consistency in effect size estimation ³⁴.

The meta-analysis was performed by utilizing correlation coefficients (r) as the primary effect size measure. Due to variations in sample sizes, study methodologies, and effect sizes, a random-effects model was applied to account for between- study heterogeneity ^{17, 35}. To assess heterogeneity, Cochran's Q test and the I² statistic were computed, with an I² value exceeding 50% indicating moderate-to-high heterogeneity ²⁴. Publication bias was assessed using Egger's regression test and funnel plot analysis, where asymmetry suggested possible small-study effects ²⁸.

Since this study relied exclusively on publicly available peer-reviewed data, no direct human subjects were involved, and Institutional Review Board (IRB) approval was not required. This meta-analysis adheres to rigorous standards of transparency and reproducibility, with clearly defined inclusion criteria, systematic selection, and robust statistical analysis. The findings contribute valuable insights into the role of engagement in academic achievement among nontraditional students in community colleges, informing educational research and policy development ^{24, 25}.

3. Result

A random-effects meta-analysis was conducted to estimate the relationship between student engagement and mathematics or academic achievement among nontraditional students. Using Fisher r-to-z transformed correlation coefficients as the outcome metric, four studies (k = 4) were included in the analysis. The overall estimated effect size was r = 0.422, with a 95% confidence interval (CI) ranging from 0.113 to 0.731. This result was statistically significant (Z = 2.68, p = .007), indicating a moderate and positive association between engagement and academic achievement. However, substantial heterogeneity was observed across studies, as indicated by the Q-test for heterogeneity, Q(3) = 50.27, p < .001, and an I² value of 92.86%, suggesting that nearly 93% of the variability in effect sizes was due to true heterogeneity rather than sampling error ³⁶. The estimated between-study variance (tau²) was 0.0888. The 95% prediction interval ranged from -0.239 to 1.083, implying that although the average effect was positive, individual studies may report considerably different findings, including potential negative effects. Diagnostic checks revealed no evidence of statistical outliers or overly influential studies, as assessed via studentized residuals and Cook's distances. Furthermore, publication bias was not detected; the fail-safe N was 1,392 (p < .001), while Begg and Mazumdar's rank correlation test (p = .750) and Egger's regression test (p = .540) were non-significant. The trim-and-fill method did not impute any missing studies. Collectively, these findings suggest that engagement significantly contributes to academic performance in nontraditional students, although contextual and methodological differences across studies warrant further investigation.

4. Discussion

The results of this meta-analysis reveal a moderate and statistically significant positive association between student engagement and mathematics or academic achievement among nontraditional students (r = 0.42, p = .007). This finding reinforces the substantial role of engagement – defined across behavioral, cognitive, and emotional domains – in shaping academic success for learners facing unique challenges, such as balancing education with employment, caregiving, or financial constraints. These results align with prior literature that emphasizes engagement as a key determinant of student outcomes across educational settings ^{1, 2}.

However, a key methodological nuance in this meta-analysis is the composition of the included studies. Among the four studies reviewed, only one ³⁷ specifically focused on mathematics achievement, using mathematics self-efficacy-a variable closely linked to cognitive and emotional engagement-as a predictor. The other three studies examined general academic achievement, primarily operationalized through Grade Point Average (GPA), which may encompass performance across a variety of subjects beyond mathematics. While this discrepancy may appear limiting, existing literature supports the validity of GPA as a general proxy for domain-specific achievement in meta-analytic contexts ^{22, 23}, especially when no sufficient mathematics-specific data are available. Moreover, self-efficacy in mathematics has been consistently linked to all three dimensions of engagement, suggesting conceptual alignment with the meta-analysis framework ^{18, 19}.

The substantial heterogeneity observed in the analysis (I² = 92.86%) suggests that the strength of the engagement-achievement relationship varies considerably across contexts. Several factors could explain this variation, including differences in measurement tools, institutional types, sample sizes, and engagement dimensions assessed. For instance, the largest study (N = 6,281) reported a strong positive correlation between academic engagement and GPA, potentially driving the overall effect upward. In contrast, one smaller study (N = 113) found a negligible effect, possibly reflecting differences in engagement measurement or student characteristics.

Another source of heterogeneity could be the multi-dimensional nature of engagement itself. Article 4 ³⁸, the only study to incorporate behavioral, emotional, and cognitive engagement simultaneously, reported a modest effect size (r = 0.24), suggesting that composite engagement may interact differently with academic outcomes compared to unidimensional measures. Additionally, the unique circumstances of nontraditional students-such as age, work commitments, and caregiving responsibilities—may moderate how engagement influences performance, particularly in mathematics, which is often a source of anxiety and disengagement for this demographic ^{12, 13}.

This meta-analysis offers important implications on student engagement and mathematics or academic achievement among nontraditional students. Notably, while only one article directly focused on mathematics achievement ³⁷, its findings offer valuable insights into the connection between engagement and achievement in mathematics contexts. Hawk's study found that mathematics self-efficacy – a construct closely associated with both cognitive and emotional engagement – significantly predicted math grades. This suggests that enhancing nontraditional students' confidence in their mathematical abilities can directly impact mathematics achievement especially. Nonetheless, the limited number of studies targeting mathematics achievement specifically signals a significant gap in literature.

Marquardt ³⁹, who utilized NSSE data from a large national sample, confirmed that quality of interactions was the strongest predictor of GPA among nontraditional students. While student-faculty interaction and effective teaching were also examined, it was the quality of the interpersonal environment – peer and faculty support – that significantly influenced academic achievement. This emphasizes the importance of fostering meaningful institutional relationships to maintain engagement, particularly for students who may be balancing work and caregiving responsibilities.

Young ³⁸ contributed a multidimensional view by analyzing behavioral, emotional, and cognitive engagement collectively. Although the effect size was modest (r = 0.24), this study supports the notion that engagement is not a monolithic construct. Rather, different types of engagement may influence achievement in distinct ways. For instance, behavioral engagement (like class participation and effort) may not have the same impact as cognitive strategies such as self-regulated learning.

Arjomandi ⁴⁰ added nuance by revealing that while nontraditional students reported higher overall engagement levels than traditional peers, active teaching strategies had only a weak link to their academic outcomes. Interestingly, nontraditional students reported greater gains in personal and professional skills, suggesting that while academic achievement may not always improve, engagement still contributes meaningfully to broader educational outcomes.

These insights collectively underscore that while engagement has a meaningful impact on mathematics achievement or academic achievement in general, the pathways through which this occurs may differ based on how engagement is defined and the contextual variables surrounding the learners ^{1, 2}. The diversity of engagement types and the unique characteristics of nontraditional students highlight the need for tailored pedagogical approaches that go beyond traditional academic metrics ⁴¹. More research is needed to isolate and evaluate the unique contribution of mathematics-specific engagement strategies for this population.

This meta-analysis is not without limitations. First, the small number of eligible studies (k = 4) limits the generalizability of findings and constrains moderator analysis. Second, while GPA served as a proxy for mathematics achievement in most studies, it cannot fully capture mathematics-specific performance nuances. Third, engagement constructs were inconsistently defined and measured, complicating direct comparisons across studies.

Future research should aim to explicitly examine the relationship between mathematics engagement and mathematics achievement among nontraditional students, using validated multidimensional engagement instruments and mathematics-specific outcome measures. Longitudinal and mixed-methods approaches may also illuminate how engagement evolves over time and interacts with external responsibilities unique to nontraditional students.

This meta-analysis provides evidence of a moderate and statistically significant relationship between student engagement and academic achievement among nontraditional students. Although only one of the four included studies explicitly focused on mathematics achievement, the remaining studies used general academic indicators such as GPA, which, while broader in scope, are accepted proxies in the absence of discipline-specific data. The inclusion of math self-efficacy in one study further reinforces the theoretical link between engagement – particularly cognitive and emotional dimensions – and mathematics achievement.

The findings underscore the relevance of fostering student engagement as a pathway to academic success in nontraditional learners. Given the unique challenges faced by this population – such as work obligations, family responsibilities, and prior academic gaps – supporting behavioral, cognitive, and emotional engagement can serve as a vital strategy to enhance learning outcomes, particularly in subjects like mathematics that often pose significant hurdles.

However, the high heterogeneity observed across studies signals the need for more precise, discipline-specific research. Future investigations should explicitly examine how engagement relates to mathematics achievement among nontraditional students using validated engagement measures and mathematics-specific outcomes. Such research is crucial for designing targeted interventions and instructional practices that respond to the needs of this growing and diverse student population.

References